Big Data Classification via Mathematical Programming

PI: Antonio Manuel Rodríguez Chía

Acronym: BIDCLAMP

Abstract: The main outcome of this project will be a new family of algorithms, models, tools, and technologies for optimizing the classification of discriminant systems in data intensive applications paying particular attention to develop flexible models and feature selection in Support Vector Machines. Our proposal is based on a deep analysis from the methodological and modeling point of view of these problems using Mathematical Programming approaches, including Linear Programming (LP), Mixed-Integer Linear Programming (MILP), Nonlinear Programming and Support Vector Machines (SVM). As a first step, the applications will focus on classification healthcare data analysis for medical diagnosis (the allocation of patients to disease classes based on symptoms and lab tests).